Evaluation of simulation tools for expansion planning
MetadataShow full item record
This master thesis is performed at the Department of Electric Power Engineering at the Norwegian University of Science and Technology in Trondheim in the spring 2008. The thesis is done in collaboration with Statkraft Energi AS and is a continuation of a specialization project from autumn 2007. The purpose of the master thesis is to perform an evaluation of simulation tools for expansion planning in hydro power systems. The aim is to evaluate the interaction between the EOPS model for long- and medium term scheduling and the SHOP model for short term hydro generation optimization. The SHOP model should be tested with seasonal, monthly and weekly time horizons, just to determine the optimal horizon. The goal is to consider if these tools made for use in hydro dominated power systems can interact in order to determine the market value of MW capacity and to quantify it by use of the simulation tools. Specified watercourses with different complexity are chosen in order to determine the possibilities of cooperation between the tools, and to verify their disposal of the hydro resources and the impact of the income for the particular cases. The importance of having reliable and sufficient simulation tools for decision making related to production optimization and expansion planning after the restructuring of the Norwegian power market are examined. The increasing amount of connections to continental markets has impact of the daily and seasonal price fluctuations, and it s thereby of significant value to have simulation tools with high degree of effectiveness and credibility. This is of really importance due to expansion planning in order to have sufficient capacity available for contribution at peak-prices. The watercourses and power stations which are studied have been of different structure and complexity which has resulted in interesting outcome. The aim was to compare the interaction between the models for three watercourses and hydro power systems just to examine the possible increased benefit. The watercourses was Rana which is quite simple in its structure, while Ulla-Førre is of a more complex category and at least the planned watercourse in Devoll, Albania, which is a run-off-river system in another power market. These cases have several restrictions and constraints such as minimum release of water and many small reservoirs which again lead to less storage possibilities. The price profiles which are used are different for the cases but reflect the year with nearly average inflow. There are used price data from Nord Pool which is adjusted to correspond to the price estimate of the EMPS model for the actual year. The results from the simulations show that it is considerable differences between the stochastically and deterministic models. The SHOP horizon which gives the most realistic results is the weekly horizon. The outcome of the three cases shows that the reservoir disposals are fairly similar when there are large and well regulated reservoirs in the watercourses. There are some differences in their disposal of the reservoirs and the production of the power plants when we are dealing with small reservoirs. There have been some difficulties for the SHOP model to hit the disposal from the EOPS model in the end of the week since there are many reservoirs involved, each with restrictions to fulfill. There is in some cases beneficial to use SHOP in interaction with the EOPS model for the optimization process in order to utilize the resources in the best way. But there are some challenges when it comes to the model consistence, especially the PQ (power and discharge) curves versus the detailed turbine curves. There are difficulties to hit the disposal from the EOPS model when there are many small reservoirs because of this inconsistency. It is observed small increase in the income when the watercourse are well regulated and when the marked price are non-volatile. There are discovered up to 15% increased income in systems which are coupled to markets with larger price variations, such as the Devoll case. This means that it seems beneficial to use SHOP in such cases where there are volatile prices.